No Such Thing As a Database Improvement

Unlike Will Rogers, I don’t think I’ve ever met a database upgrade I’ve liked.  I’ve been trying to figure out why.  Partly, I’ll admit I don’t like change.  I learn to use databases, use them well, and then, certainly without my consent, they are changed.  I’ve never been queried about the changes, nor on any committee that instituted the changes.  I’m not even sure end users are even consulted in the world of database improvements.  I think they are mostly business decisions made my managers and executed by computer programmers.

It seems to me that database improvement is just a code word for dumbing down a system so that more of the “average” searchers can use and benefit from the system.  And that means more customers and higher revenues.  In the 20 years since computer databases came into their own, I’ve experienced this let down time after time.

Lexis-Nexis and Westlaw were really the first two modern search databases of any consequence.  They even started off with stand alone units before desktop and laptop computers caught on.  Their searching, and specifically Lexis’s searching was great.  Then came windows and the mouse.  Eventually free language searching.  Each time, these things impeded what was a wonderful system.  You used to be able to string search your search (multiple commands all at once, which you typed out) so you could cut down your time in the database, because you were being charged not only for your searches but by the minute of just being on the database.  Now that these two legal databases are so huge with information, in order to find appropriate cases you need to field search, which is what we were doing “back in the day.”  Those databases have come full circle.

Harvard’s online card catalog, HOLLIS, was at first a unique computer creation just for Harvard.  That is, it was not adapted from the two big OPAC computer programs of the day (Ex Libris and Innovative Interfaces).  It was the best search system ever.  There was nothing you couldn’t find.  And as a librarian there was a public and private view (HULPR) and you could find when journals were checked in and all kinds of useful information as a reference librarian.  However, again, the web took over, and eventually Harvard had to go to a system that had greater support than the unique HOLLIS system.  Again, with millions of items in the database, you need robust searching techniques or you tend to miss things.  Nothing was better than using KSH (keyword in the subject heading) and knowing librarian subject words like handbooks, bibliography, biography, etc.  I find the web-based HOLLIS of today clunky.

When Ancestry.com upgraded their search system to include some algorithm for approximate searches, it was terrible.  I don’t know any genealogist who didn’t opt to stay with the “old” Ancestry search model.  I guess so many genealogists did that the “new” search option was dropped.  Thank goodness. 

Now, there is Americanancestors.org, the new website for the New England Historic Genealogical Society.  I’ve been a member of NEHGS for 25 years and using their webpage since day one.  Sadly, I have to say, it’s another “improvement” gone wrong.  The first thing is that I have to login every time.  The old system remembered my computer (even Ancestry does that).  So, it’s a wicked annoyance.  [I double checked that it wasn’t my Safari web browser.  It also happens on Firefox.] I tend to use the NEHGS databases to check information more than look up information.  With that in mind, I was distressed to see that the database for The American Genealogist, didn’t note the year coverage up front.  Eventually I know they’ll have all of TAG, but they don’t now.  I finally figured out that if you click on volume, there is a pull down menu that only goes to volume 43.  So, that sort of answered the question.  But the pull down menu is cumbersome for the Register which has over 160 volumes.  I preferred typing in the volume number --much quicker.  

It’s much harder to search a single database.  I did some searches of Mass. VRs 1841-1910 and despite choosing just that database, I got Maine marriages too.  I’ll have to explore more possibilities, but it seems that the functionality I held in high regard is lost and replaced by a more sweeping search functionality.  That’s fine, but it holds to my theory that the improvements are made to help drive more customers to the database and up revenues.  I just don’t know why those of us loyal and long-time customers always have to suffer.

Comments

[viagra online]

This is going to be a great deal of fun, which is one of the things I love about both shows - the sense of playfulness with the genre. Seems like they're going to make good use of Fargo's history of working with AIs with personality issues, a nifty choice to hang a plot around. Looking forward to it!